Some technical questions about VLF and Coils

[kt315]

the balancing is needed in ABSOLIUTELY PRICISELY tuning by the tuning loop. I wrote about it MANY TIMES. if you do not KNOW HOW TO DO so do not
do it! this is not for you. do then PI with monocoil. then VLF is not FOR YOU. I give the video AGAIN with demonstration of HOW tuning loop drift effects on the balance.
the tiny shift gives huge shift of phase and amplitude. so HOW you will turn a coil if you do not wishes to know the details? this job is fantastically amazing
PRECISE AND SURE NOT for the dummies.

[kt315]

[Qiaozhi]

the balancing is needed in ABSOLIUTELY PRICISELY tuning by the tuning loop. I wrote about it MANY TIMES. if you do not KNOW HOW TO DO so do not
do it! this is not for you. do then PI with monocoil. then VLF is not FOR YOU. I give the video AGAIN with demonstration of HOW tuning loop drift effects on the balance.
the tiny shift gives huge shift of phase and amplitude. so HOW you will turn a coil if you do not wishes to know the details? this job is fantastically amazing
PRECISE AND SURE NOT for the dummies.

KT315 - I am not disagreeing with you.
On the contrary, as I said: "... a tuning winding is definitely recommended." In fact, it is essential for a concentric coil, as shown in your video.
By the way, your original post (#14) was referring to a DD coil.

[kt315]

George, the physics of the processes is SAME. no differencies is there DD, omega, big-foot or planar coil. if you want to share your video
with DD balancing I will okhotno (sorry, can not translate in english this word) look it. no problem. I promise you after that I will write to minelab
informing all their DD coils are ONLY bll sht. ML wastes resourses on that everybody is able to make at kitchen table!

[golfnut]

The preamp out. (With a lot more static signal it tends to clip and saturate on large targets. )


The reason ~350mV or less is ok, is the level of the leaked signal from tx to rx coil doesnt change. It is fixed due to coils being potted and rigid.


Leaked signal entering the synch detectors are of a fixed size, then (as fixed) there is no rate of change of signal level - so at the high gain differentiators output their is no signal output - no beep. The reason is the differentiators can only amplify signals whose amplitude is changing. Moreover they can only amplify signals whose amplitude is changing at a specific rate.

Another reason coil signal leak not big deal, with synch dets , they are a special form of mixer - the leaked signal is exactly the same freq and has no phase changes (if Rx not co tuned)- so signal mixes to dc. or a dc offset, this is blocked by a cap and doesnt progress.


The average sweep speed of an adult, the shaft length / strength of the user - depicts where the differentiators have most gain - and its about 8Hz - to match the resultant signals from the sweep over a target - giving approx 8Hz signal.

Garret 250 is a little lower maybe 6Hz and narrow. Sweep lively on a 250 and you get nothing!. Raise the freq of the diffs with a smaller cap and swing with a bit of pace and clean up.

[fluxx]

Ok, so a quick update with some of the observations regarding coil-making and what I've learned. It may come useful to somebody. If I am wrong about anything, feel free to correct me.

1. The exact inductance is not that important if it's close to the needed inductance. However, resonant frequency is. So, tuning the capacitors (for both RX and TX) are actually important.
2. The resistance of TX should be as low as possible. RX resistance is less important, but if you can, then keep it as low as possible.
3. The null of the signal is less important than actual stability of the coil. The coils must ABSOLUTELY be as rigid as possible. No wobbly, bendable material is good for a foundation. Also, gluing is absolutely essential. If there's even a possibility of bending, it will cause problems and false alarms. I've tried hot glue gun, it is ok, but it is not perfect, and probably won't give superb results. I have not tried it yet (I will, in the future), but I would assume the best way to stabilize the coil, is to use the epoxy resin.
4. If you use hot glue gun or even epoxy resin, they produce heat and then cool as they set. Therefore, they will most definitely make your null drift away as they cure. With a hot glue gun, you can make corrections as you go, but with resin, you can't. Therefore, you might want to leave a couple of loops of RX out of the glue, so that you can fine-tune the null later, and then glue the loop. Another way I've heard people do it with resin is that they don't pour all of it, but rather do it in a few layers, and correct the null in between the layers.
5. The aluminum foil works for a shield, but I didn't yet compare it with other shields such as graphite spray or thin wire wrapped around coils (I will try the latter one, I think it may be good).
6. Make sure to also affix the cables you use, so they don't cause strain on coil wires or make them move when the cable is moving. This may cause various problems.
7. You can use a soundcard for an oscilloscope if you don't have one, but you have to use a voltage divider in order to not overload the input (it may damage the card). I used 8.2k and 100k resistors as the divider and managed to null the coil and also match the phase as per golfnut's drawings. Also, as the sample rate of the soundcard is so low, you can't really use it to view other waveforms other than simple sine signals up to about 10khz. The square waves and other signals don't really look right, they may even look like a sine, which may be misleading.

That's it for now, I will post some pictures as soon as I finish my IDX-PRO enclosure.

Also a question regarding IDX-PRO - can I use a Piezo speaker, or a normal speaker. If both, what is better?

Cheers and thank you for your help.

[Silver Dollar]

I have a coil that I hot glued and it moves too much so I wouldn't count on it being stable enough.

I think the RX winding wants to be higher resistance up to the noise level on the input op amp.
This allows a better impedance match and less noise.

The null is better if it isn't perfect but off by 40mv or there abouts. I try for perfect and the inevitable
shift will hopefully keep it in the ball park.

The speaker can make a huge difference. I found one from an old CB that gave me 1 - 2" better performance.
I've never tried the peizo but would not expect it to work as well (as tinny speakers do not work so well)...

[Davor]

Actually, with nowadays op amps, you are hardly ever going to match the coil resistance with the op amp noise, as the appropriately low noise op amps are rare and expensive. 1nV/sqrt(Hz) is equivalent to ~60ohm, and more common 4nV/sqrt(Hz) is equivalent to ~1kohm. In terms of noise performance, it is difficult to build a coil that could ruin the noise performance of nowadays low noise front ends.